Accurate simulation of biomolecular systems requires the consideration of solvation effects. The arrangement and dynamics of water close to a solute are strongly influenced by the solute itself. However, as the solute-solvent distance increases, the water properties tend to those of the bulk liquid. This suggests that bulk regions can be treated at a coarse grained (CG) level, while keeping the atomistic details around the solute. Since water represents about 80% of any biological system, this approach may offer a significant reduction in the computational cost of simulations without compromising atomistic details. We show here that mixing the popular SPC water model with a CG model for solvation (called WatFour) can effectively mimic the hydration, structure, and dynamics of molecular systems composed of pure water, simple electrolyte solutions, and solvated macromolecules. As a nontrivial example, we present simulations of the SNARE membrane fusion complex, a trimeric protein-protein complex embedded in a double phospholipid bilayer. Comparison with a fully atomistic reference simulation illustrates the equivalence between both approaches.